论文部分内容阅读
结合微粒群算法(PSO)具有执行速度快、受问题维数变化影响小的优点及支持向量机算法(SVM)结构风险最小化原理,构建了基于离散二进制微粒群(BPSO)与支持向量机的耕地驱动力因子选择方法,使用特征子集中确定的特征来训练支持向量回归机,用适应度函数来评价回归机的性能,指导BPSO的搜索。实验表明,该方法能有效地提取出耕地驱动因子的特征子集,从而降低了指标的维数,保留了关键信息,以获得知识的最小表达。
Combined with the advantages of PSO (Fast Particle Swarm Optimization) algorithm, which has the advantages of fast execution speed, small influence by the change of the dimension of the problem, and minimization of the structure risk of Support Vector Machine (SVM) algorithm, the PSO (Discrete Particle Swarm Optimization) The driving force factor selection method of cultivated land, using the features identified in the subset of features to train support vector regression machine, using the fitness function to evaluate the performance of the regression machine to guide the BPSO search. Experiments show that this method can effectively extract the characteristic subset of cultivated land driving factors, so as to reduce the dimension of the index and retain the key information so as to obtain the minimum expression of knowledge.